Predicting the effects of estrogen replacement therapy on lumbar bone mineral density in oophorectomized women: analysis of a 10-year longitudinal study

The purpose of this study was to investigate whether it is possible to predict the long-term effects of estrogen replacement therapy (ERT) on lumbar bone mineral density (BMD) of oophorectomized women based on changes in BMD. In this study, we retrospectively investigated the changes in lumbar BMD of 70 oophorectomized women under ERT for more than 10 years, and examined whether it was possible in the early stage of ERT to predict the amount of lumbar BMD based on various parameters. Seventy oophorectomized Japanese women (56.8±3.9 years old) treated with conjugated equine estrogen (oral) at a dosage of 0.625 mg/day for 10 years were enrolled. Lumbar (L2–L4) BMD was measured annually by dual-energy X-ray absorptiometry (DXA; CV<1.0%). The correlation between changes in BMD after 10 years on ERT (BMD10) and several clinical factors was examined using a stepwise multiple regression model. The change in BMD after 1 year on ERT (%BMD1) was the only independent factor that correlated with changes in BMD after 10 years on ERT; the coefficient of correlation was R ²=0.557 (R=0.750, P<0.001). Based on the %BMD1, the 70 women were divided into two groups: women with a positive change in the 1st year (%BMD10%) in group A (n=40) and those with a negative value in the first year ( %BMD1<0%) in group B (n=30). We investigated the sensitivity and specificity in the coincidence of %BMD1 changes in BMD after 10 years on ERT. The %BMD1 coincided with changes in BMD after 10 years on ERT; the sensitivity was 92.5% and specificity was 70.0%. In conclusion, changes in lumbar BMD on ERT can be predicted from the changes in lumbar BMD at the end of the 1st year.